1. Field of the Invention
The present invention relates to fuel injection systems for internal combustion engines and, more specifically, to a control system for a two-stroke engine.
2. Prior Art
Several alternatives for the pressurized air utilized in the injection are known; a separate air pump may be utilized, the air source may be derived from the cylinder of the engine during the compression or the expansion stroke, or the air may be derived from the crankcase pumping of the engine. In low cost applications it is desired to utilize the air source from the crankcase or the cylinder, so as to avoid the added cost and complexity of the separate air pump. In the application of pneumatic injection to larger cylinder sized engines, in general larger than 50 cc displacement, it is generally desirable to utilize injection pressure derived from the cylinder pressure because a high gas pressure may be obtained for injection. In smaller engines this tapping utilizes a disproportionate quantity of the cylinder charge gases and, thus, adversely affects the performance of the engine. It is therefore more practical to utilize the crankcase pumping source in such cases.
It is most beneficial to inject the fuel into the cylinder near to or slightly after the bottom dead center position of the piston. This injection timing avoids introducing the fuel into the early phase of the cylinder scavenging, and thus avoiding short circuit loss to the exhaust. Further, the fuel is introduced into the cylinder when the pressure in the cylinder in near atmospheric pressure, allowing the best use of the limited injection pressure to spray and therefore atomize the fuel charge. Thus, it is desirable to have a pneumatic injection timing near to the bottom dead center timing of the piston and that this timing be relatively constant with changing engine operational parameters such as speed and throttle position or load.
Several methods for operating an injection valve are taught in the prior art. U.S. Pat. No. 4,693,224 teaches the use of an electronic solenoid to operate the injection valve. This is generally unacceptable for application to small high speed engines because of the necessity of an engine control unit to operate the valve and the relatively high power requirement to drive the high speed solenoid, both adding prohibitive costs to the engine. The most common method of operating the valve as taught by the prior art is the use of some form of kinematic valve linkage driven from the crank shaft of the engine. These valves take the form of oscillating valves driven by cams as taught by a system called "PRO-JECT" described in an article "Pro-Ject Air-Assisted Fuel Injection System For Two-Stroke Engines", SAE 940397 from Universita di Pisa and a system from L'Institut Francais du Petrole described in an article "A New Two-Stroke Engine With Compressed Air Assisted Fuel Injection For High Efficiency Low Emissions Applications" by Duret et al. in SAE 880176, or rotating type valves as taught by Honda in an article "An Experimental Study of Stratified Scavenging Activated Radical Combustion Engine" by Ishibashi, SAE 972077. A problem exists with all the forms of kinematically driven valves in that they need precision surfaces and high quality materials for both the sealing members of the valve and the running portions of the drive. Valves mounted such that they are exposed to combustion gases must also be fashioned from expensive heat resistant materials. Additionally, many parts require lubrication which is not presently available in the simple two-stroke engine. Thus, the mechanical type valve arrangements add significant costs and complexity to the construction of the engine. Therefore, it is desirable to fashion an injection control valve that may be made of inexpensive materials and need not be manufactured to high tolerance, the valve and drive mechanism most preferably would require no high temperature capability or additional lubrication.
Further, an additional problem is commonly known to exist in the application of oscillating valves to high speed engines. The problem is that of the greatly increasing drive force required as the engine speed increases. For a fixed valve opening amplitude or lift, the acceleration required of the valve increases in proportion to the square of the valve opening frequency and therefore the engine speed. Further, the force required to drive the valve increases in proportion to the acceleration. Thus, the force required to drive the valve increases in proportion to the square of the engine speed. For single acting valve trains, that is valves actively driven in only one direction, these high drive forces lead to the use of large return springs to over come the valve inertial forces and prevent valve float, and consequently even more elevated drive forces. It is desirable to drive the valve in both directions, both open and closed, to avoid the use of large spring members and the associated high forces, while still attaining high speed operation. Mechanical means can be applied to drive the valve in both directions, however, this requires an even higher degree of precision and leads to even greater cost and complexity of the engine.
The final method of driving the injection valve is to operate the valve pneumatically. Pneumatic operation is affected by driving a piston through the use of a differential gas pressure across the two opposing faces of the piston. This piston in turn drives the valve. The use of pneumatic operation is common practice in gas flow control, in such devices as flow regulators and flow control valves such as spool valves. In engine operation pneumatically controlled valves are commonly utilized in carburetor operation for flow control, regulation of pressures and various operations such as driving liquid injections and opening addition flow paths. Examples of such use are shown in U.S. Pat. Nos. 5,377,637; 5,353,754; 5,197,417; 5,197,418; 4,846,119 and 4,813,391. In their application to engines where limited motion is required the piston is often in the form of a diaphragm, acting as the piston seal, and diaphragm plates functioning as the drive piston.
The use of pneumatic valve operation for control of a pneumatic injection system is taught in WO 96/07817 and EP 0789138A1. These systems utilize an injection valve placed in the head of the combustion chamber and operated on by pressures derived from various locations of the engine to influence the valve motion.
IAPAC direct fuel injection systems which use a cam to control introduction of scavenged compressed air from a crankcase have been used in the past to reduce pollutant emissions and fuel consumption in two-stroke engines. European Patent Office patent publication No. EP 0789138 discloses a camless IAPAC system (now known as SCIP) which uses a diaphragm connected to a valve, a spring, pressure from the engine crankcase, and pressure from combustion expansion gases in the combustion chamber to delay movement of the valve.
A problem exists with the cam driven IAPAC system in that added components increase cost to the engine. A problem exists with the SCIP system in that misfires in the combustion chamber result in no combustion expansion gases to delay movement of the valve. Misfires in a two-stroke engine can happen as often as one out of every three piston cycles. Thus, injection of fuel and air into the combustion chamber using a SCIP system can result in a substantial number of premature injections; about one-third of the time.
WO 96/07817 teaches a pneumatic valve that is opened when the injection pressure as derived from the crankcase of the engine overcomes the pressure from the valve closing spring and a delayed pressure wave derived from the crankcase. A problem exists in such a system that the injection pressure as derived from the crankcase is highly dependent on the engine operating condition. The peak pressure attained by the crankcase in a small two stroke engine varies with the throttle position. At wide open throttle (WOT) the peak pressure may reach 6 to 7 pounds per square inch above atmospheric pressure (psig), while at low throttle opening the peak pressure only reaches 1.5 to 2 psig. Thus the injection pressure available to open the valve is highly dependent on operating condition and thus, the injection timing is dependent on operating conditions. Further, in a small high speed engine the area of the valve is severely limited by the available space in the engine. This small area and the relatively low injection pressure available to act on that area lead to a small available force for valve opening. This coupled with the previously mention phenomenon of the required high force at high speed severely limit the use in the small high speed application. Thus it is desirable to have a valve actuation system that is largely independent of injection pressure, further it is desired that the primary motive force be derived from the diaphragm or drive piston such that the valve operation is largely independent of valve area.
A further problem exists with WO 96/07817. The wave used to control the injection is derived from the crankcase pressure through a long `delay` line. The delay line is used to control the time of arrival of the pressure wave at the valve. The Transit time in seconds of the pressure wave is fairly constant, however the transit and arrival timing in terms of crankshaft position, and therefore piston position, is highly dependent on engine speed. Thus, the injection timing is highly dependent on engine speed. Further the delay line also acts to attenuate the pressure wave, this attenuation is more acute with increasing engine speed. The attenuation coupled with the relatively weak crankcase wave render an inadequate control pressure in high speed/high load operation. It is desired to fashion a valve control system that is largely independent of engine speed.
Other embodiments of the art teach the use of controlling crank `cheeks` and additional delay lines to further control the pressure waves. These controlling cheeks must be made as precision valve surfaces to control the small flows associated with the valve control and thus add significant cost to the engine. The additional delay lines impart further speed dependence on the injection timing.
These deficiencies in WO 96/07817 are also pointed out in EP 0789138A1. EP 0789139A1 teaches the use of a valve as in the previous patent where the wave utilized to delay the injection is derived from the cylinder expansion gases. The expansion wave is again delivered to the valve control diaphragm through a delay line. In some embodiments the opening force available is enhanced by the use of longer delay lines from either the cylinder expansion gases or the crankcase wave and is delivered to the opposite side of the actuating diaphragm. Although this embodiment does enhance the opening force and improve on the problem of low pressure of the crankcase wave, the deficiency of the injection timing being highly dependent on engine speed is further introduced. Thus the injection behavior may only be optimized for a specific engine speed.
A further and critical problem is introduced through the use of the expansion gases to control the valve motion. Small two-stroke engines mostly exhibit poor combustion characteristics with misfire or partial combustion occurring every couple of strokes. During misfire there are no combustion expansion gases to be utilized to delay the injection. Further, due to ring seal leakage, the pressure during the normal expansion stroke after misfire is often sub-atmospheric, thus further advancing the injection timing. Therefore, as often as every third stroke the injection occurs at, or before, the beginning of the fresh air scavenging of the cylinder, thereby short circuiting both the unburned charge from the misfired stroke and a large portion of the early injected charge for the following stroke. It is therefore desirable to fashion an injection control system that is largely independent of combustion expansion gases from combustion of an individual piston cycle.
In both of the aforementioned publications the primary motive force for the closure of the valve is a spring positioned in the diaphragm chamber. This spring must be of sufficiently low force to allow the valve to be opened by the low injection pressures or diaphragm drive forces available. This low force combined with the increasing inertial forces of the valve at high speed lead to later and later valve closure and eventually valve float. Again it is desirable to fashion a double acting valve drive system that drives the valve both open and closed in a positive way.
A normal feature of small two-stroke engines is the lack of a separate lubrication system. The lubricant is commonly delivered to the crankcase components and the piston-cylinder unit through being mixed with the fuel. In direct injected engines, including pneumatically injected engine, of the prior art the fuel with no lubricant is delivered to the combustion chamber. This requires the addition of a separate lubrication supply pump and system for the crankcase and piston-cylinder unit, thus adding cost and complexity to the engine. It is therefore desirable to have the injection system supply a limited but significant quantity of fuel oil mixture to the crankcase to meet the engine lubrication requirement with limited additional complexity or cost.
SAE Paper 941678 entitled "Delayed Charging: A Means to Improve Two-Stoke Engine Characteristics" by Rochelle and SAE Paper 951784 entitled "Emission and Fuel Consumption Reduction in a Two-stroke Engine Using Delayed-Charging" by Rochelle, disclose use of an attempted constant flow pressure, by use of a surge tank, but this promotes leaking of fuel between the piston and the cylinder that can increase hydrocardon emissions. Rochelle also has a physically open path between the crankcase and the combustion chamber throught the surge tank at one point which the present invention avoids. Rochelle also neglects acoustic effects.